On Wed, 5 Jan 2005 15:15:01 -0500
Gene Heskett <gene.heskett AT verizon DOT net> wrote:


> >Replacing the above with one entry per filesystem (i.e wd0a, wd0e,
> > wd0g) where the whole filesystem was needed, and the top level
> > directory (/var) for the other case, with an exclude file to
> > eliminate the unwanted had the whole set dumping correctly.  I have
> > no idea if this is a generic Amanda issue or one specific to the
> > OpenBSD port.
> >
> >Debugging was complicated by the disk entries being tried in reverse
> >order, something else that does not seem to be mentioned in the
> >documentation.
> >
> >In case anyone wonders about the effect of 'inparrallel' I left it
> > at the default of 4.
> 
> Did you perchance increase the etimeout and dtimeout values first?
> 

Nope, partly as there is either no documentation on them or it's well
hidden.

I'm not sure it would have had any effect anyway as the original problem
situation had amandad failing very quickly (less than 5 seconds, I never
managed to catch it running with ps) with status 1.


> I have had as high as 53 entries for a single client in my disklist 
> without any problems.  As long as the spindle numbers used are 
> assigned to the individual drive per number, I've not even had any 
> disk thrashing problems either.
> 
> 

This was all one spindle.